_yaml: Add 'as_str()' on ScalarNode and 'get_scalar()' on MappingNode

- 'get_scalar()' allows retrieving a scalar node from a mapping.
- 'as_str()' casts a ScalarNode into a string (thus removing the node
  information).

Both together, those replace 'node_get(mapping, key, type=str)' but
also allow retrieving the 'Node' itself, which will allow in the future
lazier provenance computation.

25 files changed, 111 insertions, 79 deletions

diff --git a/src/buildstream/_cas/casremote.py b/src/buildstream/_cas/casremote.py
index cd46e9c38..0195e217a 100644
--- a/src/buildstream/_cas/casremote.py
+++ b/src/buildstream/_cas/casremote.py
@@ -32,24 +32,24 @@ class CASRemoteSpec(namedtuple('CASRemoteSpec', 'url push server_cert client_key
     @staticmethod
     def _new_from_config_node(spec_node, basedir=None):
         _yaml.node_validate(spec_node, ['url', 'push', 'server-cert', 'client-key', 'client-cert', 'instance-name'])
-        url = _yaml.node_get(spec_node, str, 'url')
+        url = spec_node.get_str('url')
         push = _yaml.node_get(spec_node, bool, 'push', default_value=False)
         if not url:
             provenance = _yaml.node_get_provenance(spec_node, 'url')
             raise LoadError(LoadErrorReason.INVALID_DATA,
                             "{}: empty artifact cache URL".format(provenance))
 
-        instance_name = _yaml.node_get(spec_node, str, 'instance-name', default_value=None)
+        instance_name = spec_node.get_str('instance-name', default=None)
 
-        server_cert = _yaml.node_get(spec_node, str, 'server-cert', default_value=None)
+        server_cert = spec_node.get_str('server-cert', default=None)
         if server_cert and basedir:
             server_cert = os.path.join(basedir, server_cert)
 
-        client_key = _yaml.node_get(spec_node, str, 'client-key', default_value=None)
+        client_key = spec_node.get_str('client-key', default=None)
         if client_key and basedir:
             client_key = os.path.join(basedir, client_key)
 
-        client_cert = _yaml.node_get(spec_node, str, 'client-cert', default_value=None)
+        client_cert = spec_node.get_str('client-cert', default=None)
         if client_cert and basedir:
             client_cert = os.path.join(basedir, client_cert)
 
diff --git a/src/buildstream/_context.py b/src/buildstream/_context.py
index c44962d7e..e285f6b6a 100644
--- a/src/buildstream/_context.py
+++ b/src/buildstream/_context.py
@@ -222,7 +222,7 @@ class Context():
             # Allow the ~ tilde expansion and any environment variables in
             # path specification in the config files.
             #
-            path = _yaml.node_get(defaults, str, directory)
+            path = defaults.get_str(directory)
             path = os.path.expanduser(path)
             path = os.path.expandvars(path)
             path = os.path.normpath(path)
@@ -254,7 +254,7 @@ class Context():
         cache = defaults.get_mapping('cache')
         _yaml.node_validate(cache, ['quota', 'pull-buildtrees', 'cache-buildtrees'])
 
-        self.config_cache_quota_string = _yaml.node_get(cache, str, 'quota')
+        self.config_cache_quota_string = cache.get_str('quota')
         try:
             self.config_cache_quota = utils._parse_size(self.config_cache_quota_string,
                                                         self.casdir)
@@ -291,8 +291,8 @@ class Context():
         self.log_verbose = _yaml.node_get(logging, bool, 'verbose')
         self.log_error_lines = _yaml.node_get(logging, int, 'error-lines')
         self.log_message_lines = _yaml.node_get(logging, int, 'message-lines')
-        self.log_element_format = _yaml.node_get(logging, str, 'element-format')
-        self.log_message_format = _yaml.node_get(logging, str, 'message-format')
+        self.log_element_format = logging.get_str('element-format')
+        self.log_message_format = logging.get_str('message-format')
 
         # Load scheduler config
         scheduler = defaults.get_mapping('scheduler')
@@ -756,7 +756,7 @@ class Context():
 #    LoadError, when the value is not of the expected type, or is not found.
 #
 def _node_get_option_str(node, key, allowed_options):
-    result = _yaml.node_get(node, str, key)
+    result = node.get_str(key)
     if result not in allowed_options:
         provenance = _yaml.node_get_provenance(node, key)
         raise LoadError(LoadErrorReason.INVALID_DATA,
diff --git a/src/buildstream/_frontend/cli.py b/src/buildstream/_frontend/cli.py
index 53adb188c..f2f149403 100644
--- a/src/buildstream/_frontend/cli.py
+++ b/src/buildstream/_frontend/cli.py
@@ -100,7 +100,7 @@ def complete_target(args, incomplete):
             return []
 
     # The project is not required to have an element-path
-    element_directory = _yaml.node_get(project, str, 'element-path', default_value='')
+    element_directory = project.get_str('element-path', default='')
 
     # If a project was loaded, use its element-path to
     # adjust our completion's base directory
diff --git a/src/buildstream/_gitsourcebase.py b/src/buildstream/_gitsourcebase.py
index 8c6b54e2d..eefa36b1d 100644
--- a/src/buildstream/_gitsourcebase.py
+++ b/src/buildstream/_gitsourcebase.py
@@ -376,7 +376,7 @@ class _GitSourceBase(Source):
     BST_MIRROR_CLASS = _GitMirror
 
     def configure(self, node):
-        ref = self.node_get_member(node, str, 'ref', None)
+        ref = node.get_str('ref', None)
 
         config_keys = ['url', 'track', 'ref', 'submodules',
                        'checkout-submodules', 'ref-format',
@@ -390,11 +390,11 @@ class _GitSourceBase(Source):
         tags = self._load_tags(node)
         self.track_tags = self.node_get_member(node, bool, 'track-tags', False)
 
-        self.original_url = self.node_get_member(node, str, 'url')
+        self.original_url = node.get_str('url')
         self.mirror = self.BST_MIRROR_CLASS(self, '', self.original_url, ref, tags=tags, primary=True)
-        self.tracking = self.node_get_member(node, str, 'track', None)
+        self.tracking = node.get_str('track', None)
 
-        self.ref_format = self.node_get_member(node, str, 'ref-format', 'sha1')
+        self.ref_format = node.get_str('ref-format', 'sha1')
         if self.ref_format not in ['sha1', 'git-describe']:
             provenance = self.node_provenance(node, member_name='ref-format')
             raise SourceError("{}: Unexpected value for ref-format: {}".format(provenance, self.ref_format))
@@ -415,7 +415,7 @@ class _GitSourceBase(Source):
         modules = node.get_mapping('submodules', {})
         for path, _ in self.node_items(modules):
             submodule = modules.get_mapping(path)
-            url = self.node_get_member(submodule, str, 'url', None)
+            url = submodule.get_str('url', None)
 
             # Make sure to mark all URLs that are specified in the configuration
             if url:
@@ -464,7 +464,7 @@ class _GitSourceBase(Source):
         return Consistency.INCONSISTENT
 
     def load_ref(self, node):
-        self.mirror.ref = self.node_get_member(node, str, 'ref', None)
+        self.mirror.ref = node.get_str('ref', None)
         self.mirror.tags = self._load_tags(node)
 
     def get_ref(self):
@@ -665,8 +665,8 @@ class _GitSourceBase(Source):
         tags = []
         tags_node = self.node_get_member(node, list, 'tags', [])
         for tag_node in tags_node:
-            tag = self.node_get_member(tag_node, str, 'tag')
-            commit_ref = self.node_get_member(tag_node, str, 'commit')
+            tag = tag_node.get_str('tag')
+            commit_ref = tag_node.get_str('commit')
             annotated = self.node_get_member(tag_node, bool, 'annotated')
             tags.append((tag, commit_ref, annotated))
         return tags
diff --git a/src/buildstream/_loader/loader.py b/src/buildstream/_loader/loader.py
index 6036f4f07..4ac598faf 100644
--- a/src/buildstream/_loader/loader.py
+++ b/src/buildstream/_loader/loader.py
@@ -227,7 +227,7 @@ class Loader():
                                 message, detail=detail) from e
             else:
                 raise
-        kind = _yaml.node_get(node, str, Symbol.KIND)
+        kind = node.get_str(Symbol.KIND)
         if kind == "junction":
             self._first_pass_options.process_node(node)
         else:
@@ -306,7 +306,7 @@ class Loader():
                         dep_deps = extract_depends_from_node(dep_element.node)
                         loader_queue.append((dep_element, list(reversed(dep_deps)), []))
 
-                        if _yaml.node_get(dep_element.node, str, Symbol.KIND) == 'junction':
+                        if dep_element.node.get_str(Symbol.KIND) == 'junction':
                             raise LoadError(LoadErrorReason.INVALID_DATA,
                                             "{}: Cannot depend on junction"
                                             .format(dep.provenance))
@@ -467,17 +467,17 @@ class Loader():
         meta_sources = []
 
         sources = _yaml.node_get(node, list, Symbol.SOURCES, default_value=[])
-        element_kind = _yaml.node_get(node, str, Symbol.KIND)
+        element_kind = node.get_str(Symbol.KIND)
 
         # Safe loop calling into _yaml.node_get() for each element ensures
         # we have good error reporting
         for i in range(len(sources)):
             source = _yaml.node_get(node, dict, Symbol.SOURCES, indices=[i])
-            kind = _yaml.node_get(source, str, Symbol.KIND)
+            kind = source.get_str(Symbol.KIND)
             _yaml.node_del(source, Symbol.KIND)
 
             # Directory is optional
-            directory = _yaml.node_get(source, str, Symbol.DIRECTORY, default_value=None)
+            directory = source.get_str(Symbol.DIRECTORY, default=None)
             if directory:
                 _yaml.node_del(source, Symbol.DIRECTORY)
 
diff --git a/src/buildstream/_loader/types.pyx b/src/buildstream/_loader/types.pyx
index c100b77b2..a3e49b040 100644
--- a/src/buildstream/_loader/types.pyx
+++ b/src/buildstream/_loader/types.pyx
@@ -85,7 +85,7 @@ cdef class Dependency:
                 _yaml.node_validate(<_yaml.Node> dep, ['filename', 'type', 'junction'])
 
                 # Make type optional, for this we set it to None
-                dep_type = <str> _yaml.node_get(<_yaml.Node> dep, str, <str> Symbol.TYPE, None, None)
+                dep_type = (<_yaml.MappingNode> dep).get_str(<str> Symbol.TYPE, None)
                 if dep_type is None or dep_type == <str> Symbol.ALL:
                     dep_type = None
                 elif dep_type not in [Symbol.BUILD, Symbol.RUNTIME]:
@@ -94,9 +94,9 @@ cdef class Dependency:
                                     "{}: Dependency type '{}' is not 'build', 'runtime' or 'all'"
                                     .format(provenance, dep_type))
 
-            self.name = <str> _yaml.node_get(<_yaml.Node> dep, str, <str> Symbol.FILENAME)
+            self.name = (<_yaml.MappingNode> dep).get_str(<str> Symbol.FILENAME)
             self.dep_type = dep_type
-            self.junction = <str> _yaml.node_get(<_yaml.Node> dep, str, <str> Symbol.JUNCTION, None, None)
+            self.junction = (<_yaml.MappingNode> dep).get_str(<str> Symbol.JUNCTION, None)
 
         else:
             raise LoadError(LoadErrorReason.INVALID_DATA,
diff --git a/src/buildstream/_options/option.py b/src/buildstream/_options/option.py
index ffdb4d272..6aa4cdc01 100644
--- a/src/buildstream/_options/option.py
+++ b/src/buildstream/_options/option.py
@@ -60,8 +60,9 @@ class Option():
     #    node (dict): The loaded YAML dictionary describing
     #                 the option
     def load(self, node):
-        self.description = _yaml.node_get(node, str, 'description')
-        self.variable = _yaml.node_get(node, str, 'variable', default_value=None)
+        self.description = node.get_str('description')
+        variable_node = node.get_str('variable', default=None)
+        self.variable = variable_node
 
         # Assert valid symbol name for variable name
         if self.variable is not None:
diff --git a/src/buildstream/_options/optionbool.py b/src/buildstream/_options/optionbool.py
index bdbb1d32a..caeff9711 100644
--- a/src/buildstream/_options/optionbool.py
+++ b/src/buildstream/_options/optionbool.py
@@ -38,7 +38,7 @@ class OptionBool(Option):
 
     def load_value(self, node, *, transform=None):
         if transform:
-            self.set_value(transform(_yaml.node_get(node, str, self.name)))
+            self.set_value(transform(node.get_str(self.name)))
         else:
             self.value = _yaml.node_get(node, bool, self.name)
 
diff --git a/src/buildstream/_options/optionenum.py b/src/buildstream/_options/optionenum.py
index 889db965c..04a19395b 100644
--- a/src/buildstream/_options/optionenum.py
+++ b/src/buildstream/_options/optionenum.py
@@ -56,7 +56,7 @@ class OptionEnum(Option):
         self.value = self.load_default_value(node)
 
     def load_value(self, node, *, transform=None):
-        self.value = _yaml.node_get(node, str, self.name)
+        self.value = node.get_str(self.name)
         if transform:
             self.value = transform(self.value)
         self.validate(self.value, _yaml.node_get_provenance(node, self.name))
@@ -79,6 +79,6 @@ class OptionEnum(Option):
                             "Valid values: {}".format(", ".join(self.values)))
 
     def load_default_value(self, node):
-        value = _yaml.node_get(node, str, 'default')
+        value = node.get_str('default')
         self.validate(value, _yaml.node_get_provenance(node, 'default'))
         return value
diff --git a/src/buildstream/_options/optionpool.py b/src/buildstream/_options/optionpool.py
index de3af3e15..25b96fa98 100644
--- a/src/buildstream/_options/optionpool.py
+++ b/src/buildstream/_options/optionpool.py
@@ -71,7 +71,7 @@ class OptionPool():
             p = _yaml.node_get_provenance(options, option_name)
             _yaml.assert_symbol_name(p, option_name, "option name", allow_dashes=False)
 
-            opt_type_name = _yaml.node_get(option_definition, str, 'type')
+            opt_type_name = option_definition.get_str('type')
             try:
                 opt_type = _OPTION_TYPES[opt_type_name]
             except KeyError:
@@ -249,7 +249,7 @@ class OptionPool():
     #
     def _process_one_node(self, node):
         conditions = _yaml.node_get(node, list, '(?)', default_value=None)
-        assertion = _yaml.node_get(node, str, '(!)', default_value=None)
+        assertion = node.get_str('(!)', default=None)
 
         # Process assersions first, we want to abort on the first encountered
         # assertion in a given dictionary, and not lose an assertion due to
diff --git a/src/buildstream/_plugincontext.py b/src/buildstream/_plugincontext.py
index 7a5407cf6..c6b6526cb 100644
--- a/src/buildstream/_plugincontext.py
+++ b/src/buildstream/_plugincontext.py
@@ -141,16 +141,16 @@ class PluginContext():
                 if kind not in _yaml.node_get(origin, list, 'plugins'):
                     continue
 
-                if _yaml.node_get(origin, str, 'origin') == 'local':
-                    local_path = _yaml.node_get(origin, str, 'path')
+                if origin.get_str('origin') == 'local':
+                    local_path = origin.get_str('path')
                     source = self._get_local_plugin_source(local_path)
-                elif _yaml.node_get(origin, str, 'origin') == 'pip':
-                    package_name = _yaml.node_get(origin, str, 'package-name')
+                elif origin.get_str('origin') == 'pip':
+                    package_name = origin.get_str('package-name')
                     source, defaults = self._get_pip_plugin_source(package_name, kind)
                 else:
                     raise PluginError("Failed to load plugin '{}': "
                                       "Unexpected plugin origin '{}'"
-                                      .format(kind, _yaml.node_get(origin, str, 'origin')))
+                                      .format(kind, origin.get_str('origin')))
                 loaded_dependency = True
                 break
 
diff --git a/src/buildstream/_project.py b/src/buildstream/_project.py
index e23689d99..bb1b06b43 100644
--- a/src/buildstream/_project.py
+++ b/src/buildstream/_project.py
@@ -84,7 +84,7 @@ class ProjectConfig:
         self.source_overrides = {}               # Source specific configurations
         self.mirrors = OrderedDict()             # contains dicts of alias-mappings to URIs.
         self.default_mirror = None               # The name of the preferred mirror.
-        self._aliases = {}                       # Aliases dictionary
+        self._aliases = None                     # Aliases dictionary
 
 
 # Project()
@@ -200,7 +200,7 @@ class Project():
 
         if url and utils._ALIAS_SEPARATOR in url:
             url_alias, url_body = url.split(utils._ALIAS_SEPARATOR, 1)
-            alias_url = _yaml.node_get(config._aliases, str, url_alias, default_value=None)
+            alias_url = config._aliases.get_str(url_alias, default=None)
             if alias_url:
                 url = alias_url + url_body
 
@@ -264,7 +264,7 @@ class Project():
     #
     def get_path_from_node(self, node, key, *,
                            check_is_file=False, check_is_dir=False):
-        path_str = _yaml.node_get(node, str, key)
+        path_str = node.get_str(key)
         path = Path(path_str)
         full_path = self._absolute_directory_path / path
 
@@ -404,7 +404,7 @@ class Project():
         else:
             config = self.config
 
-        return _yaml.node_get(config._aliases, str, alias, default_value=None)
+        return config._aliases.get_str(alias, default=None)
 
     # get_alias_uris()
     #
@@ -419,7 +419,7 @@ class Project():
         else:
             config = self.config
 
-        if not alias or alias not in config._aliases:
+        if not alias or alias not in config._aliases:  # pylint: disable=unsupported-membership-test
             return [None]
 
         mirror_list = []
@@ -429,7 +429,7 @@ class Project():
                     mirror_list = alias_mapping[alias] + mirror_list
                 else:
                     mirror_list += alias_mapping[alias]
-        mirror_list.append(_yaml.node_get(config._aliases, str, alias))
+        mirror_list.append(config._aliases.get_str(alias))
         return mirror_list
 
     # load_elements()
@@ -585,7 +585,7 @@ class Project():
 
         # The project name, element path and option declarations
         # are constant and cannot be overridden by option conditional statements
-        self.name = _yaml.node_get(self._project_conf, str, 'name')
+        self.name = self._project_conf.get_str('name')
 
         # Validate that project name is a valid symbol name
         _yaml.assert_symbol_name(_yaml.node_get_provenance(pre_config_node, 'name'),
@@ -621,7 +621,7 @@ class Project():
                         ignore_unknown=True)
 
         # Use separate file for storing source references
-        self.ref_storage = _yaml.node_get(pre_config_node, str, 'ref-storage')
+        self.ref_storage = pre_config_node.get_str('ref-storage')
         if self.ref_storage not in [ProjectRefStorage.INLINE, ProjectRefStorage.PROJECT_REFS]:
             p = _yaml.node_get_provenance(pre_config_node, 'ref-storage')
             raise LoadError(LoadErrorReason.INVALID_DATA,
@@ -715,7 +715,7 @@ class Project():
         # Perform environment expansion right away
         shell_environment = shell_options.get_mapping('environment', default={})
         for key in _yaml.node_keys(shell_environment):
-            value = _yaml.node_get(shell_environment, str, key)
+            value = shell_environment.get_str(key)
             self._shell_environment[key] = os.path.expandvars(value)
 
         # Host files is parsed as a list for convenience
@@ -730,8 +730,8 @@ class Project():
                 _yaml.node_validate(host_file_desc, ['path', 'host_path', 'optional'])
 
                 # Parse the host mount
-                path = _yaml.node_get(host_file_desc, str, 'path')
-                host_path = _yaml.node_get(host_file_desc, str, 'host_path', default_value=None)
+                path = host_file_desc.get_str('path')
+                host_path = host_file_desc.get_str('host_path', default=None)
                 optional = _yaml.node_get(host_file_desc, bool, 'optional', default_value=False)
                 mount = HostMount(path, host_path, optional)
 
@@ -806,8 +806,8 @@ class Project():
         output.options.export_variables(output.base_variables)
 
         # Override default_mirror if not set by command-line
-        output.default_mirror = self._default_mirror or _yaml.node_get(overrides, str,
-                                                                       'default-mirror', default_value=None)
+        output.default_mirror = self._default_mirror or overrides.get_str(
+            'default-mirror', default=None)
 
         mirrors = _yaml.node_get(config, list, 'mirrors', default_value=[])
         for mirror in mirrors:
@@ -815,7 +815,7 @@ class Project():
                 'name', 'aliases'
             ]
             _yaml.node_validate(mirror, allowed_mirror_fields)
-            mirror_name = _yaml.node_get(mirror, str, 'name')
+            mirror_name = mirror.get_str('name')
             alias_mappings = {}
             for alias_mapping, uris in _yaml.node_items(mirror.get_mapping('aliases')):
                 assert isinstance(uris, list)
@@ -880,7 +880,7 @@ class Project():
             allowed_origins = ['core', 'local', 'pip']
             _yaml.node_validate(origin, allowed_origin_fields)
 
-            origin_value = _yaml.node_get(origin, str, 'origin')
+            origin_value = origin.get_str('origin')
             if origin_value not in allowed_origins:
                 raise LoadError(
                     LoadErrorReason.INVALID_YAML,
@@ -907,7 +907,7 @@ class Project():
 
             # Store the origins if they're not 'core'.
             # core elements are loaded by default, so storing is unnecessary.
-            if _yaml.node_get(origin, str, 'origin') != 'core':
+            if origin.get_str('origin') != 'core':
                 self._store_origin(origin, 'sources', plugin_source_origins)
                 self._store_origin(origin, 'elements', plugin_element_origins)
 
@@ -947,7 +947,7 @@ class Project():
                 if group in origin_node:
                     _yaml.node_del(origin_node, group)
 
-            if _yaml.node_get(origin_node, str, 'origin') == 'local':
+            if origin_node.get_str('origin') == 'local':
                 path = self.get_path_from_node(origin, 'path',
                                                check_is_dir=True)
                 # paths are passed in relative to the project, but must be absolute
diff --git a/src/buildstream/_variables.pyx b/src/buildstream/_variables.pyx
index 9b8b5a902..c41ff5c3f 100644
--- a/src/buildstream/_variables.pyx
+++ b/src/buildstream/_variables.pyx
@@ -128,7 +128,7 @@ cdef class Variables:
         cdef str value
 
         for key in _yaml.node_keys(node):
-            value = <str> _yaml.node_get(node, str, key)
+            value = node.get_str(key)
             ret[sys.intern(key)] = _parse_expstr(value)
         return ret
 
diff --git a/src/buildstream/_workspaces.py b/src/buildstream/_workspaces.py
index 1adc52ac2..a4f7abc76 100644
--- a/src/buildstream/_workspaces.py
+++ b/src/buildstream/_workspaces.py
@@ -631,8 +631,8 @@ class Workspaces():
     def _load_workspace(self, node):
         dictionary = {
             'prepared': _yaml.node_get(node, bool, 'prepared', default_value=False),
-            'path': _yaml.node_get(node, str, 'path'),
-            'last_successful': _yaml.node_get(node, str, 'last_successful', default_value=None),
+            'path': node.get_str('path'),
+            'last_successful': node.get_str('last_successful', default=None),
             'running_files': _yaml.node_sanitize(
                 node.get_mapping('running_files', default=None),
                 dict_type=dict),
diff --git a/src/buildstream/_yaml.pxd b/src/buildstream/_yaml.pxd
index 7d7e87d8d..b2237f075 100644
--- a/src/buildstream/_yaml.pxd
+++ b/src/buildstream/_yaml.pxd
@@ -31,6 +31,12 @@ cdef class Node:
 cdef class MappingNode(Node):
     cdef Node get(self, str key, default, default_constructor)
     cpdef MappingNode get_mapping(self, str key, default=*)
+    cpdef ScalarNode get_scalar(self, str key, default=*)
+    cpdef str get_str(self, str key, object default=*)
+
+
+cdef class ScalarNode(Node):
+    cpdef str as_str(self)
 
 
 cdef class ProvenanceInformation:
diff --git a/src/buildstream/_yaml.pyx b/src/buildstream/_yaml.pyx
index 9d9a12400..a7779c41e 100644
--- a/src/buildstream/_yaml.pyx
+++ b/src/buildstream/_yaml.pyx
@@ -78,11 +78,19 @@ cdef class Node:
 cdef class ScalarNode(Node):
 
     def __init__(self, object value, int file_index, int line, int column):
+        if type(value) is str:
+            value = value.strip()
         self.value = value
         self.file_index = file_index
         self.line = line
         self.column = column
 
+    cpdef str as_str(self):
+        # We keep 'None' as 'None' to simplify the API's usage and allow chaining for users
+        if self.value is None:
+            return None
+        return str(self.value)
+
 
 cdef class MappingNode(Node):
 
@@ -119,6 +127,25 @@ cdef class MappingNode(Node):
 
         return value
 
+    cpdef ScalarNode get_scalar(self, str key, object default=_sentinel):
+        value = self.get(key, default, ScalarNode)
+
+        if type(value) is not ScalarNode:
+            if value is None:
+                value = ScalarNode(None, _SYNTHETIC_FILE_INDEX, 0, next_synthetic_counter())
+            else:
+                provenance = node_get_provenance(value)
+                raise LoadError(LoadErrorReason.INVALID_DATA,
+                                "{}: Value of '{}' is not of the expected type 'Scalar'"
+                                .format(provenance, key))
+
+        return value
+
+    cpdef str get_str(self, str key, object default=_sentinel):
+        # TODO: don't go through 'get_scalar', we can directly get everything and optimize
+        cdef ScalarNode scalar = self.get_scalar(key, default)
+        return scalar.as_str()
+
 
 class SequenceNode(Node):
     def __init__(self, list value, int file_index, int line, int column):
diff --git a/src/buildstream/element.py b/src/buildstream/element.py
index aaf8577a5..05e590823 100644
--- a/src/buildstream/element.py
+++ b/src/buildstream/element.py
@@ -512,7 +512,7 @@ class Element(Plugin):
           # variables in the returned string
           name = self.node_subst_member(node, 'name')
         """
-        value = self.node_get_member(node, str, member_name, default)
+        value = node.get_str(member_name, default)
         try:
             return self.__variables.subst(value)
         except LoadError as e:
@@ -2702,7 +2702,7 @@ class Element(Plugin):
         # Sandbox config, unlike others, has fixed members so we should validate them
         _yaml.node_validate(sandbox_config, ['build-uid', 'build-gid', 'build-os', 'build-arch'])
 
-        build_arch = _yaml.node_get(sandbox_config, str, 'build-arch', default_value=None)
+        build_arch = sandbox_config.get_str('build-arch', default=None)
         if build_arch:
             build_arch = Platform.canonicalize_arch(build_arch)
         else:
@@ -2711,7 +2711,7 @@ class Element(Plugin):
         return SandboxConfig(
             _yaml.node_get(sandbox_config, int, 'build-uid'),
             _yaml.node_get(sandbox_config, int, 'build-gid'),
-            _yaml.node_get(sandbox_config, str, 'build-os', default_value=host_os),
+            sandbox_config.get_str('build-os', default=host_os),
             build_arch)
 
     # This makes a special exception for the split rules, which
diff --git a/src/buildstream/plugins/elements/junction.py b/src/buildstream/plugins/elements/junction.py
index 849df72be..d30e34828 100644
--- a/src/buildstream/plugins/elements/junction.py
+++ b/src/buildstream/plugins/elements/junction.py
@@ -175,9 +175,9 @@ class JunctionElement(Element):
     BST_FORBID_RDEPENDS = True
 
     def configure(self, node):
-        self.path = self.node_get_member(node, str, 'path', default='')
+        self.path = node.get_str('path', default='')
         self.options = node.get_mapping('options', default={})
-        self.target = self.node_get_member(node, str, 'target', default=None)
+        self.target = node.get_str('target', default=None)
         self.target_element = None
         self.target_junction = None
 
diff --git a/src/buildstream/plugins/sources/_downloadablefilesource.py b/src/buildstream/plugins/sources/_downloadablefilesource.py
index b9b15e268..7b1785979 100644
--- a/src/buildstream/plugins/sources/_downloadablefilesource.py
+++ b/src/buildstream/plugins/sources/_downloadablefilesource.py
@@ -77,8 +77,8 @@ class DownloadableFileSource(Source):
     __urlopener = None
 
     def configure(self, node):
-        self.original_url = self.node_get_member(node, str, 'url')
-        self.ref = self.node_get_member(node, str, 'ref', None)
+        self.original_url = node.get_str('url')
+        self.ref = node.get_str('ref', None)
         self.url = self.translate_url(self.original_url)
         self._warn_deprecated_etag(node)
 
@@ -99,7 +99,7 @@ class DownloadableFileSource(Source):
             return Consistency.RESOLVED
 
     def load_ref(self, node):
-        self.ref = self.node_get_member(node, str, 'ref', None)
+        self.ref = node.get_str('ref', None)
         self._warn_deprecated_etag(node)
 
     def get_ref(self):
@@ -143,7 +143,7 @@ class DownloadableFileSource(Source):
                                   .format(self.url, sha256, self.ref))
 
     def _warn_deprecated_etag(self, node):
-        etag = self.node_get_member(node, str, 'etag', None)
+        etag = node.get_str('etag', None)
         if etag:
             provenance = self.node_provenance(node, member_name='etag')
             self.warn('{} "etag" is deprecated and ignored.'.format(provenance))
diff --git a/src/buildstream/plugins/sources/bzr.py b/src/buildstream/plugins/sources/bzr.py
index e59986da6..082d11dec 100644
--- a/src/buildstream/plugins/sources/bzr.py
+++ b/src/buildstream/plugins/sources/bzr.py
@@ -69,9 +69,9 @@ class BzrSource(Source):
     def configure(self, node):
         self.node_validate(node, ['url', 'track', 'ref', *Source.COMMON_CONFIG_KEYS])
 
-        self.original_url = self.node_get_member(node, str, 'url')
-        self.tracking = self.node_get_member(node, str, 'track')
-        self.ref = self.node_get_member(node, str, 'ref', None)
+        self.original_url = node.get_str('url')
+        self.tracking = node.get_str('track')
+        self.ref = node.get_str('ref', None)
         self.url = self.translate_url(self.original_url)
 
     def preflight(self):
diff --git a/src/buildstream/plugins/sources/deb.py b/src/buildstream/plugins/sources/deb.py
index e45994951..cc88cf53c 100644
--- a/src/buildstream/plugins/sources/deb.py
+++ b/src/buildstream/plugins/sources/deb.py
@@ -61,7 +61,7 @@ class DebSource(TarSource):
     def configure(self, node):
         super().configure(node)
 
-        self.base_dir = self.node_get_member(node, str, 'base-dir', None)
+        self.base_dir = node.get_str('base-dir', None)
 
     def preflight(self):
         return
diff --git a/src/buildstream/plugins/sources/pip.py b/src/buildstream/plugins/sources/pip.py
index 9d6c40d74..db0537497 100644
--- a/src/buildstream/plugins/sources/pip.py
+++ b/src/buildstream/plugins/sources/pip.py
@@ -111,8 +111,8 @@ class PipSource(Source):
     def configure(self, node):
         self.node_validate(node, ['url', 'packages', 'ref', 'requirements-files'] +
                            Source.COMMON_CONFIG_KEYS)
-        self.ref = self.node_get_member(node, str, 'ref', None)
-        self.original_url = self.node_get_member(node, str, 'url', _PYPI_INDEX_URL)
+        self.ref = node.get_str('ref', None)
+        self.original_url = node.get_str('url', _PYPI_INDEX_URL)
         self.index_url = self.translate_url(self.original_url)
         self.packages = self.node_get_member(node, list, 'packages', [])
         self.requirements_files = self.node_get_member(node, list, 'requirements-files', [])
diff --git a/src/buildstream/plugins/sources/remote.py b/src/buildstream/plugins/sources/remote.py
index 562a8f226..4e9fe4f7d 100644
--- a/src/buildstream/plugins/sources/remote.py
+++ b/src/buildstream/plugins/sources/remote.py
@@ -62,7 +62,7 @@ class RemoteSource(DownloadableFileSource):
     def configure(self, node):
         super().configure(node)
 
-        self.filename = self.node_get_member(node, str, 'filename', os.path.basename(self.url))
+        self.filename = node.get_str('filename', os.path.basename(self.url))
         self.executable = self.node_get_member(node, bool, 'executable', False)
 
         if os.sep in self.filename:
diff --git a/src/buildstream/plugins/sources/tar.py b/src/buildstream/plugins/sources/tar.py
index c90de74ea..ef2ee1d85 100644
--- a/src/buildstream/plugins/sources/tar.py
+++ b/src/buildstream/plugins/sources/tar.py
@@ -72,8 +72,7 @@ class TarSource(DownloadableFileSource):
     def configure(self, node):
         super().configure(node)
 
-        self.base_dir = self.node_get_member(node, str, 'base-dir', '*')
-
+        self.base_dir = node.get_str('base-dir', '*')
         self.node_validate(node, DownloadableFileSource.COMMON_CONFIG_KEYS + ['base-dir'])
 
     def preflight(self):
diff --git a/src/buildstream/plugins/sources/zip.py b/src/buildstream/plugins/sources/zip.py
index 9981f1260..1724f0587 100644
--- a/src/buildstream/plugins/sources/zip.py
+++ b/src/buildstream/plugins/sources/zip.py
@@ -72,8 +72,7 @@ class ZipSource(DownloadableFileSource):
     def configure(self, node):
         super().configure(node)
 
-        self.base_dir = self.node_get_member(node, str, 'base-dir', '*')
-
+        self.base_dir = node.get_str('base-dir', '*')
         self.node_validate(node, DownloadableFileSource.COMMON_CONFIG_KEYS + ['base-dir'])
 
     def get_unique_key(self):